Double-decision Response Time Models of Recall and Recognition Support Resource Accounts of Visual Working Memory

Area of Science:

• Cognitive Psychology
• Neuroscience
• Computational Modeling

Background:

• Visual working memory (VWM) is crucial for cognitive tasks.
• Debate exists whether VWM operates on a fixed-capacity 'slot' model or a flexible 'resource' model.

Purpose of the Study:

• To determine if VWM is better described by a slot or resource model.
• To investigate the nature of memory representations in VWM.

Main Methods:

• Utilized a double-decision paradigm combining continuous-outcome recall and two-choice recognition.
• Applied continuous-outcome and two-choice EZ diffusion models to analyze response times and accuracy.
• Estimated latent cognitive parameters: drift rate (memory strength) and boundary separation.

Main Results:

• Found significant correlations in drift rate and boundary separation between recall and recognition tasks, indicating shared processes.
• Demonstrated that 'heavy tails' in recall distributions represent low-precision memory traces, not random guesses.
• Simulations indicated swap errors minimally contributed to these heavy tails.

Conclusions:

• Results support variable precision and resource accounts of VWM, contradicting slot-based models.
• Highlights the utility of joint modeling of response time and accuracy data.
• Emphasizes the value of double-decision tasks for understanding memory processes.